sattle

A Python package for LSST alert verification.

Description

sattle is a Python package that provides alert verification for Rubin Observatory alerts, built on top of Bill Gray's satellite tracking code.

Requirements

• Python 3.11.11
• C++ compiler and build tools
• Docker/Apptainer (for containerized deployment)

Dependencies

Python packages:

• lsst-sphgeom
• lsst-utils
• aiohttp
• requests
• pybind11

External repositories:

• lunar (https://github.com/Bill-Gray/lunar)
• sat_code (https://github.com/Bill-Gray/sat_code)

Installation

The easiest way to run the package is using Docker:

# Build the Docker image
docker build -t sattle .

# Run the container
docker run -p 9999:9999 sattle

For USDF use, the container is run via apptainer, you need to bind the output to local files.

apptainer --debug run  --contain  --no-home  --pwd /output /
--env-file ~sattle/.credentials.sh /
--bind ~/apptainer_logs:/app/sattle/python/lsst/sattle/logs /
--bind ~/apptainer_output:/output  sattle_tickets-dm-51091.sif /
python /app/sattle/bin.src/app.py

Manual Installation

1. Clone the required repositories:

   git clone --branch tickets/DM-49713 https://github.com/lsst-dm/sattle.git
   git clone https://github.com/Bill-Gray/lunar.git
   git clone https://github.com/Bill-Gray/sat_code.git

2. Install Python dependencies:

   pip install lsst-sphgeom lsst-utils aiohttp requests pybind11

3. Build and install lunar:

   cd lunar
   make install

4. Build sat_code:

   cd sat_code
   make

5. Build sattle.so

sattle.so must be built within sat_code.

..code-block:: bash run c++ -O3 -Wall -shared -std=c++11 $(python3 -m pybind11 --includes)

observe.cpp sdp4.cpp sgp4.cpp sgp8.cpp sdp8.cpp sattle.cpp sgp.o deep.cpp common.cpp basics.cpp get_el.cpp -o sattle$(python3-config --extension-suffix) -fPIC

Then copy the output .so file as sattle.so into ~/sattle/python/lsst/sattle/

Usage

The package provides a server that runs on port 9999 by default. After starting the server:

python app.py

The server will be available at http://localhost:9999. You can now make api calls to calculate a cache for specific visits. Please refer to sattle/bin.src/example_client.py for example puts. The first call is made during pipe_tasks in the AP pipelines to populate the comparison catalog. The second put call is made in detectAndMeasure to verify the dia sources and return diaSource ids which will be included in the catalog.

If you are using historical data, you must include historical=True in the requests.